Integrated hydrogen production and refueling station intelligent management system

Abstract

The present invention discloses an integrated intelligent management system for hydrogen production and refueling stations, and relates to the technical field of energy management. The system includes: a historical data acquisition module, which is used to determine the target station characteristics of a target hydrogen production and refueling station, and acquire historical record data according to the target station characteristics, wherein the historical record data includes historical hydrogenation data and historical hydrogen production data; a working condition clustering module, which is used to cluster the working conditions of the historical record data, and define multiple hydrogen production working conditions, wherein the hydrogen production working conditions are associated with working condition application limits; a power model construction module, which is used to construct a multi-condition hydrogen production power model based on historical hydrogen production data and multiple hydrogen production working conditions; a decision-making and management execution module, which is used to make hydrogen production decisions based on real-time hydrogenation needs according to the multi-condition hydrogen production power model, and manage and execute the target hydrogen production and refueling station based on the decision results. Thereby achieving the technical effect of improving the operating efficiency of hydrogen production and refueling stations and reducing operating costs.

Description

Technical Field

[0001] The present invention relates to the technical field of energy management, and in particular to an integrated hydrogen production and hydrogenation station intelligent management system. Background Art

[0002] As the global demand for clean energy continues to increase, hydrogen energy, as a green energy, has huge development potential. The production and refueling of hydrogen is a crucial link in the hydrogen energy industry chain, and the management efficiency of hydrogen production and hydrogen refueling stations directly affects the operating efficiency of the entire hydrogen energy supply chain. At present, hydrogen production and hydrogen refueling stations mainly rely on manual experience and simple automatic control systems for management, which is difficult to cope with complex operating conditions and dynamically changing needs. There are technical problems such as mismatch between supply and demand, and affecting operational efficiency and energy utilization. Summary of the invention

[0003] The present invention provides an integrated intelligent management system for hydrogen production and hydrogenation stations to solve the technical problems of supply and demand mismatch in the prior art, affecting operational efficiency and energy utilization, and achieve the technical effect of improving the operational efficiency and energy utilization of hydrogen production and hydrogenation stations and reducing operating costs.

[0004] The integrated hydrogen production and hydrogenation station intelligent management system provided by the present invention comprises:

[0005] A historical data acquisition module is used to determine the target station characteristics of the target hydrogen production and hydrogenation station, and acquire historical record data based on the target station characteristics, wherein the historical record data includes historical hydrogenation data and historical hydrogen production data.

[0006] An operating condition clustering module, wherein the operating condition clustering module is used to cluster the operating conditions of the historical record data and define a plurality of hydrogen production operating conditions, wherein the hydrogen production operating conditions are associated with operating condition application limits.

[0007] A power model construction module, the power model construction module is used to construct a multi-condition hydrogen production power model based on the historical hydrogen production data and the multiple hydrogen production conditions, wherein the multi-condition hydrogen production power model includes multiple condition hydrogen production power sub-models and corresponding call constraints.

[0008] A decision-making and management execution module is used to make hydrogen production decisions based on real-time hydrogenation demand according to the multi-condition hydrogen production power model, and to manage and execute the target hydrogen production and hydrogenation station based on the decision results.

[0009] In a feasible implementation, the historical record data is clustered into operating conditions to define a plurality of hydrogen production operating conditions, wherein the hydrogen production operating conditions are associated with operating condition application limits, and the execution steps include:

[0010] Obtain the design specification information of the target hydrogen production and refueling station and extract the ideal hydrogen production parameter range.

[0011] The historical hydrogenation data is clustered to obtain a plurality of hydrogenation data clusters, wherein each of the hydrogenation data clusters corresponds to a hydrogenation operating condition.

[0012] Taking the ideal hydrogen production parameter range as a judgment constraint, traverse the plurality of hydrogenation data clusters, and call the corresponding historical hydrogen production data to perform hydrogen production adaptability judgment.

[0013] According to the hydrogen production adaptability judgment result, the historical hydrogen production data that does not meet the judgment constraint is eliminated, and based on the elimination result, a plurality of the hydrogen production conditions and the condition application limits corresponding to the plurality of the hydrogenation data clusters are generated.

[0014] In a feasible implementation, based on the historical hydrogen production data and the plurality of hydrogen production conditions, a multi-condition hydrogen production power model is constructed, and the execution steps include:

[0015] Based on the multiple hydrogen production conditions and the condition application limits, the historical hydrogen production data is remapped to obtain multiple sample condition hydrogen production data sets.

[0016] The sample quantity information of the hydrogen production data sets under the sample operating conditions is obtained by statistical analysis, and the tilt weight sets of the multiple operating conditions are defined.

[0017] The multi-operating-condition hydrogen production power model is constructed according to the tilted weight set, and supervised training of the multi-operating-condition hydrogen production power model is performed using a plurality of the sample operating-condition hydrogen production data sets.

[0018] In a feasible implementation, the multi-operating condition hydrogen production power model is constructed according to the tilted weight set, and supervised training of the multi-operating condition hydrogen production power model is performed through a plurality of the sample operating condition hydrogen production data sets, and the execution steps include:

[0019] Based on the tilt weight set, the model type and the number of models are defined, and multiple groups of hydrogen production power sub-models are constructed, and the multiple groups of hydrogen production power sub-models correspond to multiple hydrogen production conditions.

[0020] The corresponding sample operating condition hydrogen production data set is called to perform data mutation to obtain multiple sample data spaces.

[0021] Samples are extracted from the plurality of sample data spaces, and supervised training is performed by traversing a plurality of groups of hydrogen production power sub-models according to the extraction results.

[0022] The multiple groups of hydrogen production power sub-models after supervised training are integrated and learned, and a working condition index is established to obtain the multi-working condition hydrogen production power model.

[0023] In a feasible implementation, according to the multi-operating condition hydrogen production power model, a hydrogen production decision based on real-time hydrogenation demand is made, and then the execution step further includes:

[0024] A hydrogen production accounting function is defined, wherein the hydrogen production accounting function includes an energy consumption cost operator, a yield cost operator and a maintenance cost operator.

[0025] The hydrogen production accounting function is used as the objective function and combined with the corresponding operating condition hydrogen production power sub-model to perform cost optimization of the decision result.

[0026] In a feasible implementation, the cost optimization of the decision result is performed by taking the hydrogen production accounting function as the objective function and combining the corresponding working condition hydrogen production power sub-model, and the execution steps include:

[0027] The operating condition hydrogen production power sub-model is activated to perform iterative decision making and generate a decision output set.

[0028] The decision output set is traversed, the objective function values ​​of multiple decision outputs are calculated, and the optimal selection is performed according to the objective function values, and the decision result after cost optimization is output.

[0029] In a feasible implementation, before making a hydrogen production decision based on real-time hydrogenation demand according to the multi-operating condition hydrogen production power model, the execution step further includes:

[0030] A local computing power evaluation is performed on the target hydrogen production and refueling station, and a model compression coefficient set of the multi-operating condition hydrogen production power model is defined according to the computing power evaluation result.

[0031] Adaptive model compression is performed on the plurality of operating condition hydrogen production power sub-models according to the model compression coefficient set.

[0032] Based on the compressed multiple operating condition hydrogen production power sub-models, the multi-operating condition hydrogen production power model is updated.

[0033] In a feasible implementation, the target station characteristics of the target hydrogen production and hydrogenation station are determined, and historical record data is collected and acquired according to the target station characteristics. The execution steps include:

[0034] Interactively target hydrogen production and hydrogenation stations, extract hydrogen production process characteristics, hydrogenation scale characteristics, and output them as the target station characteristics.

[0035] The target station characteristics are used as call constraints to obtain homologous expanded record data of the target hydrogen production and hydrogenation station.

[0036] The homologous expanded record data is combined with the original record data of the target hydrogen production and hydrogenation station and output as the historical record data.

[0037] The present invention discloses an integrated intelligent management system for hydrogen production and hydrogenation stations, including: a historical data acquisition module, the historical data acquisition module is used to determine the target station characteristics of a target hydrogen production and hydrogenation station, and acquire historical record data according to the target station characteristics, wherein the historical record data includes historical hydrogenation data and historical hydrogen production data; an operating condition clustering module, the operating condition clustering module is used to perform operating condition clustering on the historical record data, and define multiple hydrogen production operating conditions, wherein the hydrogen production operating conditions are associated with operating condition application limits; a power model construction module, the power model construction module is used to construct a multi-operating condition hydrogen production power model based on the historical hydrogen production data and multiple hydrogen production operating conditions, wherein the multi-operating condition hydrogen production power model includes multiple operating condition hydrogen production power sub-models and corresponding call constraints; a decision-making and management execution module, the decision-making and management execution module is used to make a hydrogen production decision based on the real-time hydrogenation demand according to the multi-operating condition hydrogen production power model, and perform management execution of the target hydrogen production and hydrogenation station based on the decision result. The integrated hydrogen production and refueling station intelligent management system disclosed in the present invention solves the technical problems of supply and demand mismatch and affecting operational efficiency and energy utilization, and achieves the technical effect of improving the operational efficiency and energy utilization of hydrogen production and refueling stations and reducing operating costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 It is a structural schematic diagram of the integrated hydrogen production and hydrogenation station intelligent management system of the present invention.

[0039] Figure 2 It is a schematic diagram of a process for defining multiple hydrogen production conditions in the integrated hydrogen production and hydrogenation station intelligent management system of the present invention.

[0040] Explanation of the accompanying drawings: historical data collection module 11, working condition clustering module 12, power model construction module 13, decision-making and management execution module 14. DETAILED DESCRIPTION

[0041] The technical solution provided in the embodiments of the present invention is to solve the technical problems of mismatch between supply and demand and affecting operational efficiency and energy utilization in the prior art. The overall idea adopted is as follows:

[0042] First, a historical data acquisition module is used to determine the target station characteristics of a target hydrogen production and refueling station, and acquire historical record data based on the target station characteristics, wherein the historical record data includes historical hydrogen refueling data and historical hydrogen production data; then, a working condition clustering module is used to perform working condition clustering on the historical record data, and define multiple hydrogen production working conditions, wherein the hydrogen production working conditions are associated with working condition application limits; then, a power model construction module is used to construct a multi-working condition hydrogen production power model based on the historical hydrogen production data and multiple hydrogen production working conditions, wherein the multi-working condition hydrogen production power model includes multiple working condition hydrogen production power sub-models and corresponding call constraints; then, a decision-making and management execution module is used to make hydrogen production decisions based on real-time hydrogen refueling needs according to the multi-working condition hydrogen production power model, and perform management and execution of the target hydrogen production and refueling station based on the decision results.

[0043] The above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods of the specification to better understand the above technical solution. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments of the present invention. It should be understood that the present invention is not limited to the example embodiments used only to explain the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. In addition, it should be noted that, for the convenience of description, only the parts related to the present invention are shown in the drawings, rather than all of them.

[0044] Embodiment 1

[0045] Figure 1 The figure is a flow chart of the integrated hydrogen production and hydrogenation station intelligent management system of the present invention, wherein the system comprises:

[0046] The historical data acquisition module 11 is used to determine the target station characteristics of the target hydrogen production and hydrogenation station, and acquire historical record data according to the target station characteristics, wherein the historical record data includes historical hydrogenation data and historical hydrogen production data.

[0047] Specifically, the target station characteristics refer to the key factors or parameters that can define the hydrogen production and hydrogenation station, which are used to provide data collection targets for data collection. Exemplarily, the target station characteristics of the target hydrogen production and hydrogenation station include: hydrogen production method, hydrogen production capacity, equipment type and configuration (including equipment type, quantity and configuration), operation characteristics (load fluctuations) and environmental characteristics (climatic conditions).

[0048] Specifically, hydrogenation data refers to historical operating data related to the hydrogenation process, which is used to reflect the hydrogenation demand information of the target hydrogen production and hydrogenation station, and illustratively includes hydrogenation amount, hydrogenation pressure and flow, gas quality data, etc.; hydrogen production data is data related to the hydrogen production process, and includes information such as the hydrogen production amount, hydrogen production efficiency, and production process parameters of the hydrogen production equipment of the target hydrogen production and hydrogenation station.

[0049] By defining and analyzing the characteristics of the target hydrogen production and refueling station, historical record data can be collected and utilized. The determination of the above characteristics and the collection of data are the basis for subsequent data analysis, model training and decision support, thereby ensuring that the hydrogen production and refueling station can operate efficiently and safely.

[0050] In some embodiments, determining the target station characteristics of the target hydrogen production and hydrogenation station, and acquiring historical record data according to the target station characteristics, the execution steps include:

[0051] Interactively target hydrogen production and hydrogenation stations, extract hydrogen production process characteristics, hydrogenation scale characteristics, and output them as the target station characteristics; use the target station characteristics as call constraints to obtain the homologous extended record data of the target hydrogen production and hydrogenation station; combine the homologous extended record data with the original record data of the target hydrogen production and hydrogenation station, and output them as the historical record data.

[0052] Specifically, first, characteristic data related to the operation of the target hydrogen production and hydrogenation station is obtained, wherein the characteristics of the target station include but are not limited to hydrogen production process characteristics and hydrogenation scale characteristics. These two types of characteristics determine the production capacity, operation mode, equipment configuration, etc. of the station.

[0053] Specifically, the hydrogen production process characteristics involve key information such as the technology, equipment configuration, and energy consumption used in the hydrogen production process. The extracted hydrogen production process characteristics include hydrogen production methods, hydrogen production capacity, equipment configuration, energy consumption data, process parameter range, etc.; hydrogenation scale characteristics mainly involve the scale of hydrogenation stations, equipment configuration, and the performance of hydrogenation processes. For example, the extracted hydrogenation scale characteristics include hydrogenation pressure, hydrogenation flow, hydrogenation equipment configuration, hydrogenation power and energy consumption, gas quality control, etc. By extracting the above hydrogen production process characteristics and hydrogenation scale characteristics, the target station characteristics are finally output to provide a basis for subsequent historical record data collection.

[0054] Specifically, based on the characteristic data of the target station, the homologous data is expanded. The homologous expanded data refers to the data collected from other hydrogen production and refueling stations with similar processes, configurations and operating modes as the target station to enrich the data set of the target station.

[0055] Optionally, the main ways to expand homologous data include geographic location expansion, selecting data from other hydrogen production and refueling stations in the same or similar geographic areas for expansion; equipment type matching, collecting corresponding historical data from hydrogen refueling stations with similar equipment types and configurations; process similarity matching, selecting site data that use the same or similar hydrogen production technologies (such as water electrolysis, natural gas steam reforming, etc.); and operation and maintenance historical data matching, selecting site data with similar operating history, load fluctuations, and hydrogen refueling demands for expansion.

[0056] Furthermore, the data obtained from other similar sites will be combined with the original records of the target site to form a more complete and rich historical record data set, providing more comprehensive support for subsequent data application links.

[0057] By expanding the data, the diversity and representativeness of the data set can be enhanced, providing a wider sample set for subsequent data analysis, thereby improving the generalization ability of subsequent model training.

[0058] The operating condition clustering module 12 is used to cluster the operating conditions of the historical record data and define a plurality of hydrogen production operating conditions, wherein the hydrogen production operating conditions are associated with operating condition application limits.

[0059] Specifically, the operating conditions of historical record data are clustered, multiple hydrogen production operating conditions are defined, and the application limits of each operating condition are defined, including analysis and classification of data according to different operating conditions, environmental factors, equipment configurations and other factors in the hydrogen production process, identifying and dividing different operating scenarios, so as to better understand and optimize the hydrogen production process, where each operating condition represents a specific operating mode or working state.

[0060] In some embodiments, Figure 2 As shown, the historical record data is clustered into operating conditions to define a plurality of hydrogen production operating conditions, wherein the hydrogen production operating conditions are associated with operating condition application limits, and the execution steps include:

[0061] The design specification information of the target hydrogen production and hydrogenation station is obtained, and an ideal hydrogen production parameter range is extracted; the historical hydrogenation data is clustered to obtain a plurality of hydrogenation data clusters, wherein each of the hydrogenation data clusters corresponds to a hydrogenation operating condition; taking the ideal hydrogen production parameter range as a discrimination constraint, the plurality of hydrogenation data clusters are traversed, and the corresponding historical hydrogen production data are called to perform hydrogen production adaptability discrimination; according to the hydrogen production adaptability discrimination result, the historical hydrogen production data that does not meet the discrimination constraint is eliminated, and based on the elimination result, a plurality of hydrogen production operating conditions and the operating condition application limits corresponding to the plurality of hydrogenation data clusters are generated.

[0062] Specifically, first, the design specification information of the target hydrogen production and refueling station is obtained. The design specification information is a detailed description of the operating requirements and design parameters of the target hydrogen production and refueling station, including the ideal operating parameter range for hydrogen production and refueling, equipment operating limitations, optimized working conditions, etc. The extracted ideal hydrogen production parameter range refers to the ideal operating parameter range based on equipment and process design during the hydrogen production process, exemplarily including the reaction temperature range, reaction pressure range, mass flow rate or flow rate, which is used to provide a standard for subsequent adaptability judgment.

[0063] Specifically, the features that best reflect the changes in the working conditions during the hydrogen production process, such as hydrogen production temperature, pressure, reactant mass flow, power consumption, equipment operating status, etc., are selected as the clustering basis, and then the similarity between the historical record data is calculated based on the selected features. Exemplarily, the above similarity is obtained by calculating the Euclidean distance, cosine similarity, Manhattan distance, etc. The smaller the distance, the more similar the data points are, and the greater the possibility that they belong to the same working condition.

[0064] Specifically, according to the distribution characteristics of historical data, the number of clusters or the parameters of the clustering algorithm (such as the density threshold of DBSCAN, the K value of K-means, etc.) are preset to cluster the historical record data, and each data point is classified into a different operating condition cluster; then, multiple hydrogen production conditions are defined according to the clustering results, including naming according to the characteristics of the operating conditions, such as "high-load hydrogen production conditions", "low-temperature startup conditions", "normal operating conditions", etc.

[0065] Specifically, based on the clustering results of historical hydrogenation data, the applicability of hydrogen production conditions is screened in combination with the ideal hydrogen production parameter range, including distinguishing each hydrogenation data cluster to determine whether the elements included therein meet the ideal hydrogen production parameter range in the design specifications, thereby ensuring that each hydrogenation condition is consistent with the ideal operating parameters in the hydrogen production process, avoiding system inefficiency or equipment damage due to the mismatch between the hydrogenation process and the hydrogen production process.

[0066] Furthermore, after completing the adaptability judgment, if some historical hydrogen production data cannot meet the ideal hydrogen production conditions, these unsatisfactory data will be eliminated accordingly to ensure that the definition of subsequent operating conditions is generated based on data that can meet the ideal operating conditions, thereby ensuring the stability and efficiency of the hydrogen production process.

[0067] Specifically, each hydrogenation data cluster corresponds to a suitable hydrogen production condition after adaptive judgment. At the same time, each hydrogenation data cluster corresponds to the application scope of each condition, that is, the condition application limit, including the actual operation limits of temperature, pressure, flow, etc., so as to ensure the accuracy of the power model constructed for different conditions. For example, some hydrogenation data clusters can well characterize the hydrogen production characteristics of the target hydrogenation station under high temperature and high pressure conditions, while other clusters are suitable for low temperature and low pressure conditions.

[0068] The above process effectively identifies and defines multiple hydrogenation conditions through cluster analysis of historical hydrogenation data and adaptability judgment of ideal hydrogen production parameter ranges, and provides corresponding hydrogen production conditions and application limits for each condition, which helps to understand and optimize the hydrogenation and hydrogen production process, and then construct corresponding kinetic models for adaptability to different conditions in subsequent steps.

[0069] The power model building module 13 is used to build a multi-condition hydrogen production power model based on the historical hydrogen production data and the multiple hydrogen production conditions, wherein the multi-condition hydrogen production power model includes multiple condition hydrogen production power sub-models and corresponding call constraints.

[0070] Specifically, according to different hydrogen production conditions and based on historical hydrogen production data, the kinetic behavior of each condition in the hydrogen production process is simulated to obtain a multi-condition hydrogen production power model. The multi-condition hydrogen production power model can reflect the kinetic characteristics such as reaction rate, energy consumption, and material flow under different conditions, thereby determining the operating conditions (such as temperature, pressure, flow rate, etc.) for achieving different hydrogen production performances, and optimizing and adjusting them.

[0071] Specifically, the multi-condition hydrogen production power model includes multiple working condition hydrogen production power sub-models, each sub-model corresponds to a specific hydrogen production condition, and describes the hydrogen production process dynamics under this condition. Different working condition sub-models have different parameters and constraints, because the rate and energy efficiency of the hydrogen production reaction will vary under different working conditions. At the same time, each sub-model has corresponding calling constraints, which determine under what conditions the sub-model can be activated. For example, a certain working condition hydrogen production sub-model may only be valid within a specific temperature and pressure range, and the model cannot be used under conditions outside this range, and a different working condition sub-model needs to be used.

[0072] In some embodiments, based on the historical hydrogen production data and the plurality of hydrogen production conditions, a multi-condition hydrogen production power model is constructed, and the execution steps include:

[0073] Based on the multiple hydrogen production conditions and the condition application limits, the historical hydrogen production data is remapped to obtain multiple sample condition hydrogen production data sets; statistical analysis is performed to obtain sample quantity information of the multiple sample condition hydrogen production data sets, and tilted weight sets of multiple conditions are defined; based on the tilted weight sets, the multi-condition hydrogen production power model is constructed, and supervised training of the multi-condition hydrogen production power model is performed through the multiple sample condition hydrogen production data sets.

[0074] Specifically, first, based on the previously defined multiple hydrogen production conditions and condition application limits, the historical hydrogen production data are reallocated to the hydrogen production conditions corresponding to the condition application limits, that is, the historical data are reallocated to the correct condition data set, and the generated multiple new sample condition hydrogen production data sets represent a specific hydrogen production condition and contain the historical hydrogen production process data under the condition, so as to better represent the characteristics of the hydrogen production process under different conditions.

[0075] Then, the number of samples under each working condition is statistically analyzed, and the weight of the working condition is defined according to the imbalance of the number of samples, so as to appropriately weight the model during training and avoid the model training bias caused by the small number of samples in some working conditions. For example, for working conditions with fewer samples, the weight of the weight can be appropriately increased, indicating that the working condition needs more attention and training, so as to avoid the model being overly biased towards working conditions with more samples.

[0076] Specifically, a multi-operating condition hydrogen production power model is constructed, and the model is optimized and trained through historical data sets. The multi-operating condition hydrogen production power model is used to map hydrogen production parameters according to different hydrogen production needs.

[0077] By constructing a multi-operating condition hydrogen production dynamic model, the dynamics of the hydrogen production process under different operating conditions can be effectively described. Each operating condition corresponds to a sub-model, and multiple sub-models are managed by calling constraints to ensure that the most appropriate sub-model can be automatically selected for prediction and optimization under different operating conditions. This not only improves the flexibility and stability of the hydrogen production process, but also provides strong support for the efficient operation of hydrogen production.

[0078] In some implementations, the multi-operating-condition hydrogen production power model is constructed according to the tilted weight set, and supervised training of the multi-operating-condition hydrogen production power model is performed using a plurality of the sample operating-condition hydrogen production data sets, and the execution steps include:

[0079] Based on the tilted weight set, the model type and the number of models are defined, and a plurality of groups of hydrogen production power sub-models are constructed, wherein the plurality of groups of hydrogen production power sub-models correspond to a plurality of hydrogen production working conditions; the corresponding sample working condition hydrogen production data set is called to perform data mutation, and a plurality of sample data spaces are obtained; sample extraction is performed in the plurality of sample data spaces, and supervised training is performed by traversing the plurality of groups of hydrogen production power sub-models according to the extraction results; integrated learning of the plurality of groups of hydrogen production power sub-models after supervised training is performed, and a working condition index is established to obtain the multi-working condition hydrogen production power model.

[0080] Specifically, first, using the previously calculated tilt weight set (reflecting the deviation in the number of samples of different operating conditions), select appropriate machine learning or physical models, such as deep neural networks, support vector machines (SVM), regression models, etc., to build a multi-operating condition hydrogen production power model. Among them, operating conditions with higher tilt weights will correspond to more hydrogen production power sub-models to enhance the fitting ability of these operating conditions. For example, for operating conditions with fewer but important samples, multiple sub-models need to be integrated.

[0081] Specifically, a plurality of hydrogen production power sub-models are grouped, wherein the plurality of groups of hydrogen production power sub-models correspond to a plurality of hydrogen production working conditions.

[0082] Specifically, mutation operations are performed on each sample operating condition hydrogen production data set to increase data diversity. Increasing the diversity of training data through data mutation can enable multiple hydrogen production power sub-models to generalize better and avoid overfitting. For example, more training samples are generated through data enhancement techniques (such as noise injection, data translation, rotation, etc.). After mutation of each operating condition data set, multiple sample data spaces can be obtained, representing the performance of the same operating condition under different data mutations, thereby providing more mutation data for subsequent training.

[0083] Furthermore, from each sample data space, a sampling strategy (such as random sampling, importance sampling, weighted sampling, etc.) is adopted to extract a certain number of samples according to the inclination weight of the operating conditions, so as to ensure that the selected data can represent the data characteristics under each operating condition; then, according to the extracted sample data, each group of hydrogen production power sub-models is trained, and each sub-model in each group of hydrogen production power sub-models learns the data patterns related to specific operating conditions through the results of sample extraction with different compositions, gradually fits these data and optimizes its prediction ability.

[0084] Specifically, through an integrated learning method, multiple trained hydrogen production power sub-models are combined into a unified multi-operating condition hydrogen production power model to improve the overall prediction performance, and an operating condition index is established based on the prediction results of different operating conditions for more efficient calling. Among them, the operating condition index can find the corresponding sub-model by querying the input features of the operating condition (such as temperature, pressure, etc.) and quickly make predictions.

[0085] By integrating multiple groups of sub-models and operating condition indexes after learning, the multi-operating condition hydrogen production power model constructed can select appropriate sub-models for prediction according to the input characteristics of different operating conditions, thereby improving the prediction accuracy and efficiency of the entire model.

[0086] The decision-making and management execution module 14 is used to make a hydrogen production decision based on the real-time hydrogenation demand according to the multi-operating condition hydrogen production power model, and to manage and execute the target hydrogen production and hydrogenation station based on the decision result.

[0087] Specifically, the real-time hydrogenation demand of the target hydrogen production and refueling station is obtained in real time, and the real-time hydrogenation demand is used as the model input of the multi-condition hydrogen production power model. The multi-condition hydrogen production power model automatically identifies the current hydrogen production condition according to the input demand and parameters, and then maps and obtains the corresponding hydrogen production control parameters, and outputs the corresponding decision results to manage and execute the target hydrogen production and refueling station, wherein the real-time hydrogenation demand refers to the hydrogenation demand data of the target hydrogen production and refueling station, including the amount, time, frequency, etc. of hydrogenation.

[0088] Optionally, according to the resource allocation output in the hydrogen production decision, the equipment of the target hydrogen production and refueling station, such as electrolyzers, compressors, hydrogen storage tanks, etc., is dispatched to ensure the utilization efficiency and energy utilization of each equipment is maximized to achieve more efficient hydrogen production and supply.

[0089] Through the multi-condition hydrogen production power model, the resources required for hydrogen production are predicted based on real-time data, and through precise scheduling and optimization, on-demand supply and optimal resource allocation are achieved, ensuring that hydrogen refueling needs are met while improving the efficiency and stability of the overall system.

[0090] In some embodiments, according to the multi-condition hydrogen production power model, a hydrogen production decision based on real-time hydrogenation demand is made, and then the execution step further includes:

[0091] A hydrogen production accounting function is defined, wherein the hydrogen production accounting function includes an energy consumption cost operator, a yield cost operator and a maintenance cost operator; the hydrogen production accounting function is used as an objective function, and the cost of the decision result is optimized in combination with the corresponding operating condition hydrogen production power sub-model.

[0092] Furthermore, when making hydrogen production decisions based on real-time hydrogenation demand, in addition to considering the basic production requirements to meet hydrogenation demand, it is also necessary to further optimize costs to improve the economic benefits of the target hydrogen production and hydrogenation station. Therefore, the hydrogen production accounting function is introduced, combined with the multi-condition hydrogen production power model, to comprehensively optimize the energy consumption, yield and maintenance costs in the hydrogen production process, to ensure that the overall production cost is reduced while meeting the demand.

[0093] Specifically, a hydrogen production accounting function is constructed to evaluate various costs in the hydrogen production process, including energy consumption cost, yield cost and maintenance cost, and use them as optimization targets to guide hydrogen production decisions. The energy consumption cost operator is used to calculate the cost of energy consumption in the hydrogen production process. The energy consumption cost operator multiplies the energy consumption under different working conditions by the corresponding energy unit price to obtain the cost; the yield cost operator is used to calculate the cost caused by insufficient yield in the hydrogen production process. For example, if the hydrogen production efficiency is lower than expected, it is necessary to increase the output of the hydrogen production equipment or extend the production time, which will lead to additional operating costs. The yield cost operator performs cost evaluation based on the deviation between the hydrogen production efficiency and the target hydrogen production under different working conditions; the maintenance cost operator is used to calculate the maintenance cost. By considering the operating status of the equipment under different working conditions and its maintenance requirements, it is calculated based on factors such as the expected life of the equipment and the maintenance cycle. For example, the operating time, failure rate and maintenance frequency of the equipment will affect the maintenance cost. For different hydrogen production conditions, the maintenance cost will fluctuate accordingly.

[0094] Specifically, the hydrogen production accounting function is used as the objective function, and the multi-condition hydrogen production power sub-model is combined for optimization calculation to achieve the lowest cost hydrogen production decision. When performing cost optimization, in addition to the objective function, other constraints need to be considered, such as: hydrogenation demand satisfaction constraints, equipment capacity constraints (such as maximum production capacity, equipment life, etc.), safety and efficiency constraints.

[0095] By introducing hydrogen production accounting functions and optimization processes, and performing cost optimization based on a multi-condition hydrogen production power model, the energy consumption, yield loss, and maintenance costs in the hydrogen production process can be effectively reduced, ensuring that hydrogen production decisions can achieve the optimal cost-effectiveness while meeting demand.

[0096] In some implementations, the cost optimization of the decision result is performed by taking the hydrogen production accounting function as the objective function and combining the corresponding operating condition hydrogen production power sub-model, and the execution steps include:

[0097] Activate the working condition hydrogen production power sub-model to perform iterative decision-making and generate a decision output set; traverse the decision output set, calculate the objective function values ​​of multiple decision outputs, and make the best selection based on the objective function value, and output the decision result after cost optimization.

[0098] Optionally, first, according to the real-time hydrogen refueling demand and the target operating condition, the hydrogen production power sub-model of each operating condition is activated to perform preliminary decision calculations, wherein, after inputting the real-time hydrogen refueling demand, each operating condition hydrogen production power sub-model simulates and predicts the hydrogen production process and generates a preliminary operation strategy, that is, in each round of iteration, different hydrogen production parameters with similar hydrogen production performance are generated; then, for the decision output of each operating condition sub-model, the hydrogen production accounting function calculates its corresponding objective function value, and then selects the decision result with the smallest objective function value (that is, the optimal cost) from multiple decision outputs to achieve cost optimization.

[0099] By activating multiple working condition hydrogen production power sub-models and making iterative decisions, a series of candidate hydrogen production strategies can be generated. After that, by calculating the objective function value of each strategy and making the best selection, the decision output with the smallest objective function value is used as the final cost-optimized decision result, which not only meets the hydrogenation demand, but also achieves the optimization of the lowest energy consumption cost, yield cost and maintenance cost.

[0100] Furthermore, according to the multi-condition hydrogen production power model, a hydrogen production decision based on real-time hydrogenation demand is made, before which the execution step further includes:

[0101] Perform a local computing power assessment on the target hydrogen production and refueling station, and define a model compression coefficient set of the multi-operating condition hydrogen production power model based on the computing power assessment result; perform adaptive model compression on the multiple operating condition hydrogen production power sub-models based on the model compression coefficient set; and update the multi-operating condition hydrogen production power model based on the compressed multiple operating condition hydrogen production power sub-models.

[0102] Specifically, the current computing power resources of the target hydrogen production and refueling station are evaluated to ensure that the multi-condition hydrogen production power model is not limited by hardware performance during actual execution and can be executed efficiently within the specified time. The hardware resources of the target hydrogen production and refueling station are evaluated, including the performance of computing units such as CPU, GPU, and memory, the availability of computing resources, and factors affecting computing power such as network bandwidth, so as to determine the maximum model scale and computational complexity that the site can support under different workloads.

[0103] Optionally, the evaluation results include the processing capacity of the target hydrogen production and refueling station, the maximum number of supported model parameters, and the calculation speed, to ensure that subsequent model compression and optimization will not exceed the computing power carrying capacity.

[0104] Specifically, according to the evaluation results of local computing power, a reasonable set of model compression coefficients is defined to adaptively compress the multi-operating condition hydrogen production power model to ensure that the operating efficiency of the model is not limited by hardware. Multiple compression coefficients are compression ratios for hydrogen production power sub-models under different operating conditions. For devices with lower computing power, a higher compression coefficient is used; for devices with stronger computing power, a lower compression coefficient is used to retain the fineness of the model as much as possible.

[0105] Furthermore, according to the defined compression coefficient set, adaptive compression is performed for each operating condition hydrogen production power sub-model to reduce model complexity and ensure that the model can still run efficiently under the constraints of local computing resources. Exemplary adaptive compression methods include network pruning, quantization, low-rank decomposition, knowledge distillation, etc.

[0106] Furthermore, the adaptively compressed sub-models of hydrogen production under multiple working conditions are reintegrated to form an updated multi-condition hydrogen production model, and the overall model is retrained to ensure that the compressed sub-models can work together in hydrogen production decision-making and provide accurate prediction and decision support.

[0107] Through the above steps, the multi-condition hydrogen production power model can be adaptively compressed according to the computing power evaluation results of the target hydrogen production and refueling station to ensure effective operation under limited computing power while ensuring the accuracy and real-time nature of hydrogen production decisions.

[0108] In summary, the integrated hydrogen production and hydrogenation station intelligent management system provided by the present invention has the following technical effects:

[0109] Through the historical data collection module, the target station characteristics of the target hydrogen production and refueling station are determined, and historical record data is acquired based on the target station characteristics, wherein the historical record data includes historical hydrogen refueling data and historical hydrogen production data; the operating condition clustering module performs operating condition clustering on the historical record data and defines multiple hydrogen production operating conditions, wherein the hydrogen production operating conditions are associated with operating condition application limits; the power model construction module constructs a multi-operating condition hydrogen production power model based on the historical hydrogen production data and multiple hydrogen production operating conditions, wherein the multi-operating condition hydrogen production power model includes multiple operating condition hydrogen production power sub-models and corresponding call constraints; the decision-making and management execution module makes hydrogen production decisions based on the real-time hydrogen refueling demand according to the multi-operating condition hydrogen production power model, and manages and executes the target hydrogen production and refueling station based on the decision results. Thereby achieving the technical effect of improving the operating efficiency and energy utilization of hydrogen production and refueling stations and reducing operating costs.

[0110] It should be understood that the embodiments disclosed in the present invention and the above description can enable those skilled in the art to use the present invention to implement the present invention. At the same time, the present invention is not limited to the above-mentioned embodiments. It should be understood that those skilled in the art can still modify the technical solutions recorded in the above-mentioned embodiments, or replace some of the technical features therein by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. Integrated hydrogen production and hydrogenation station intelligent management system, characterized by: The system comprises: A historical data acquisition module, the historical data acquisition module is used to determine the target station characteristics of the target hydrogen production and hydrogenation station, and acquire historical record data according to the target station characteristics, wherein the historical record data includes historical hydrogenation data and historical hydrogen production data; A working condition clustering module, the working condition clustering module is used to cluster the historical record data into working conditions, and define a plurality of hydrogen production working conditions, wherein the hydrogen production working conditions are associated with working condition application limits, and the working condition application limits are defined based on the application range of each working condition; A power model construction module, wherein the power model construction module is used to construct a multi-condition hydrogen production power model based on the historical hydrogen production data and the plurality of hydrogen production conditions, wherein the multi-condition hydrogen production power model includes a plurality of condition hydrogen production power sub-models and corresponding call constraints; A decision-making and management execution module, which is used to make a hydrogen production decision based on the real-time hydrogenation demand according to the multi-operating condition hydrogen production power model, and to manage and execute the target hydrogen production and hydrogenation station based on the decision result; The operating condition clustering is performed on the historical record data to define multiple hydrogen production operating conditions, and the execution steps include: Obtain the design specification information of the target hydrogen production and refueling station and extract the ideal hydrogen production parameter range; Clustering the historical hydrogenation data to obtain a plurality of hydrogenation data clusters, wherein each of the hydrogenation data clusters corresponds to a hydrogenation operating condition; Taking the ideal hydrogen production parameter range as a discrimination constraint, traversing a plurality of hydrogenation data clusters, and calling the corresponding historical hydrogen production data to perform hydrogen production adaptability discrimination; According to the hydrogen production adaptability judgment result, the historical hydrogen production data that does not meet the judgment constraint is eliminated, and based on the elimination result, a plurality of the hydrogen production conditions and the condition application limits corresponding to the plurality of the hydrogenation data clusters are generated; Wherein, based on the historical hydrogen production data and the plurality of hydrogen production conditions, a multi-condition hydrogen production power model is constructed, and the execution steps include: Based on the plurality of hydrogen production operating conditions and the operating condition application limits, remapping the historical hydrogen production data is performed to obtain a plurality of sample operating condition hydrogen production data sets; Statistically analyzing and acquiring sample quantity information of a plurality of the sample operating condition hydrogen production data sets, and defining tilted weight sets of a plurality of operating conditions, wherein the tilted weight sets are defined based on the quantity distribution of the sample quantity information; The multi-operating-condition hydrogen production power model is constructed according to the tilted weight set, and supervised training of the multi-operating-condition hydrogen production power model is performed using a plurality of the sample operating-condition hydrogen production data sets.

2. The integrated hydrogen production and hydrogenation station intelligent management system according to claim 1, characterized in that: According to the tilted weight set, the multi-operating condition hydrogen production power model is constructed, and supervised training of the multi-operating condition hydrogen production power model is performed through a plurality of the sample operating condition hydrogen production data sets, and the execution steps include: Based on the tilt weight set, define the model type and the number of models, and construct multiple groups of hydrogen production power sub-models, wherein the multiple groups of hydrogen production power sub-models correspond to multiple hydrogen production conditions; Calling the corresponding sample operating condition hydrogen production data set to perform data mutation to obtain multiple sample data spaces; Extracting samples from a plurality of the sample data spaces, and traversing a plurality of groups of the hydrogen production power sub-models for supervised training according to the extraction results; The multiple groups of hydrogen production power sub-models after supervised training are integrated and learned, and a working condition index is established to obtain the multi-working condition hydrogen production power model.

3. The integrated hydrogen production and hydrogenation station intelligent management system according to claim 2, characterized in that: According to the multi-condition hydrogen production power model, a hydrogen production decision is made based on real-time hydrogenation demand, and then the execution step further includes: A hydrogen production accounting function is defined, wherein the hydrogen production accounting function includes an energy consumption cost operator, a yield cost operator and a maintenance cost operator; The hydrogen production accounting function is used as the objective function and combined with the corresponding operating condition hydrogen production power sub-model to perform cost optimization of the decision result.

4. The integrated hydrogen production and hydrogenation station intelligent management system according to claim 3, characterized in that: Taking the hydrogen production accounting function as the objective function and combining the corresponding working condition hydrogen production power sub-model to perform cost optimization of the decision result, the execution steps include: Activating the working condition hydrogen production power sub-model to perform iterative decision making and generate a decision output set; The decision output set is traversed, the objective function values ​​of multiple decision outputs are calculated, and the optimal selection is performed according to the objective function values, and the decision result after cost optimization is output.

5. The integrated hydrogen production and hydrogenation station intelligent management system according to claim 1, characterized in that: According to the multi-condition hydrogen production power model, a hydrogen production decision is made based on real-time hydrogenation demand. Before that, the execution step also includes: Performing a local computing power evaluation on the target hydrogen production and refueling station, and defining a model compression coefficient set of the multi-operating condition hydrogen production power model according to the computing power evaluation result; Performing adaptive model compression on the plurality of hydrogen production power sub-models under the operating conditions according to the model compression coefficient set; Based on the compressed multiple operating condition hydrogen production power sub-models, the multi-operating condition hydrogen production power model is updated.

6. The integrated hydrogen production and hydrogenation station intelligent management system according to claim 1, characterized in that: Determine the target station characteristics of the target hydrogen production and hydrogenation station, and collect and obtain historical record data according to the target station characteristics, and the execution steps include: Interactive target hydrogen production and hydrogenation station, extract hydrogen production process characteristics, hydrogenation scale characteristics, and output as the target station characteristics; Taking the target station characteristics as call constraints, obtaining homologous expanded record data of the target hydrogen production and hydrogenation station; The homologous expanded record data is combined with the original record data of the target hydrogen production and hydrogenation station and output as the historical record data.

Images